06-reference

notboring great blue frontier

Wed Apr 22 2026 20:00:00 GMT-0400 (Eastern Daylight Time) ·reference ·source: Not Boring ·by Packy McCormick + Will O'Brien (co-written; guest co-author is President of Ulysses Maritime)
frontier-thesisvertical-integrationinnovators-dilemmaocean-economycurve-convergence

“The Great Blue Frontier” — @packyM + @WillOBrienOcean

Why this is in the vault

Packy gives over the keyboard to a portfolio-adjacent founder pitching the ocean as the next industrial frontier — and in doing so, restates his own “Curve Convergence” thesis cleanly enough to lift it into our positioning library. The mapping is medium for RDCO: the ocean specifics are tangential, but the underlying frontier-industrialization framework (governing-metric collapse, vertical-integration playbook, Innovator’s Dilemma against incumbents) is exactly the lens we use for AI-native data work vs. legacy data consultancies.

⚠️ Sponsorship & bias

Two layers worth flagging:

The core argument

Three nested moves:

  1. Frontier-industrialization template. Frontiers (West, space) get tamed when supply-side tech and demand-side urgency converge AND someone collapses the cost of operating in the domain — railroad for the West, reusable rockets ($/kg-to-orbit) for space. The ocean has had episodic exploration but never industrialization; it still runs on an “1800s expedition model” of expensive crewed vessels on expensive campaigns.
  2. Why-now: Curve Convergence. Four exogenous curves crossed thresholds in the 2020-2025 window — connectivity (Starlink), energy (wave/floating power), edge compute/AI (data-center-class GPUs in AUV form factor), and motors/sensors (commodified via drones and EVs). Demand side: undersea cables, critical minerals, defense maritime-domain awareness.
  3. The incumbent trap is textbook Innovator’s Dilemma. Ocean primes (Lockheed, RTX, HII, Fugro, Oceaneering) are locked into cost-plus, crewed, campaign-based service models because that’s what their best customers (navies, oil majors, telcos) demand. The “New Ocean” cohort is vertically integrated, autonomy-native, software-defined, built on commercial supply chains, priced for scale — same shape as the New Space cohort vs. Lockheed/Boeing/Northrop a decade ago.

The governing metric Will proposes for the ocean is ”$/km²-month of active ocean work” — explicitly modeled on $/kg-to-orbit. Ulysses claims to be attacking it on both OpEx (autonomous integrated surface+subsea system, no crewed mothership) and CapEx (in-house manufactured AUVs at ~$50K vs. $500K-$5M incumbent).

Mapping against Ray Data Co

Three load-bearing transfers, one weak one:

Strong: the governing-metric framing. “Every constrained economy has one number that, if you move it, everything else follows.” This is exactly the discipline missing from most data-consultancy positioning. RDCO’s analog is some version of $/insight-delivered or $/decision-supported — the cost of getting a defensible, audit-trail-quality answer out of a stack. If we can articulate that number and credibly bend it (via harness-and-fat-skills automation), the frontier-industrialization rhetoric becomes a positioning weapon, not just a metaphor. Worth a Sanity Check arc.

Strong: Innovator’s Dilemma against legacy data consultancies. Will’s description of why ocean primes can’t pivot — locked-in workforce, mil-spec supplier network, every incentive pointing at the existing book of business — maps cleanly onto Big-Four data practices and traditional dbt-shop consultancies. They listen to the current market too closely. RDCO’s positioning advantage isn’t being smarter; it’s being unencumbered.

Medium: vertical integration as default for new domains. Will frames vertical integration as forced (“the existing supply chains are slow, expensive, not built for scale”). This validates RDCO’s pattern of building our own harness/skills rather than assembling vendor tooling — same root cause, same playbook. Cross-references the Hengsperger reindustrialize-America thesis we filed Apr 19.

Weak: actual ocean opportunities. Direct ocean exposure is not on RDCO’s roadmap. Don’t fabricate data-modeling-for-AUVs angles. The frontier framework is the asset, not the domain.

Notable framing devices to steal

One quote worth pulling

“The cheapest AUV in the world is still expensive if it’s hand launched.” (≤15 words; captures the autonomy-or-bust principle that applies to AI tooling too — the cheapest skill is expensive if it requires a human babysitter.)

Substack-rendered plaintext via Gmail; redirect-wrapped links. Paraphrased throughout, single ≤15-word quote in quotation marks per copy-paste caution.